Conceptexamenprogramma Maritiem en Techniek

29  Download (0)

Hele tekst


DIPARTIMENTO DI SCIENZE MEDICO-VETERINARIE Corso di Laurea Magistrale a Ciclo Unico in Medicina Veterinaria




Relatore: Prof. Federico RIGHI


Dott.ssa Marica SIMONI


Elisa BATTAGLIA Anno Accademico 2019/2020


Ad Anna, Irene e Chiara, per essere entrate nella mia vita e non essersene mai andate.



Abstract... 1

Chapter 1: Protein metabolism in the rumen ... 2

1.1 – Protein degradation ... 2

1.2 – Microbial protein synthesis... 3

Chapter 2: Rumen fermentation synchronisation... 6

2.1 – Protein and energy synchronisation ... 6

2.2 – The Cornell Net Carbohydrate and Protein System ... 7

Chapter 3: Nitrogen efficiency and protein nutrition in dairy cows ... 10

3.1 – Rumen degradable and undegradable protein ... 11

3.2 – Limiting amino acids ... 12

Chapter 4: Nitrogen efficiency and environmental pollution ... 15

4.1 – The nitrogen cascade ... 15

4.2 – Dairy industry and nitrogen pollution ... 16

Chapter 5: Casein nitrogen as an efficiency parameter ... 17

Chapter 6: Experimental contribution ... 19

Introduction ... 19

Materials and methods ... 21

Results and discussion ... 25

Conclusions ... 37

Bibliography ... 38

Website references ... 43




The objective of the present study was to evaluate the milk and casein nitrogen (N) use efficiency in farms in which cows were fed hay-based total-mixed ration (TMR).

Nineteen Italian Holstein dairy farms were randomly selected within the Parmigiano- Reggiano cheese production area.

Dry matter intake (DMI) was calculated based on the amount of TMR offered and orts weighed over a two days sampling period. The TMR was analysed to estimate dry matter (DM), neutral detergent fibre (NDF), starch, crude protein (CP) and N content, then N intake was calculated. For each farm, data on milk yield and composition were recorded; milk N efficiency (MNE) and casein N efficiency (CNE) were calculated as a ratio between milk or casein N yield and N intake. The farms were then divided in two categories based on milk yield: group 0 (≤ 33 Kg/day) and group 1 (> 33 Kg/day). The data were normally distributed, and they were analysed through the mixed model as a function of farm categories, with farm as random effect. Group 0 (average diet: 15.2% CP, 38.7% NDF and 21.5% starch) presented an average milk yield of 30.86 Kg/cow/day, while group 1 (average diet: 15.3% CP, 37.9% NDF; 22.0% starch) presented an average milk yield of 35.86 Kg/cow/day.

No differences were highlighted between diet chemical composition, nutrient intake and milk composition. A trend for a higher CP intake but lower milk fat and protein percentage was observed in group 1, even if not statistically. No differences were observed concerning MNE and CNE between the two groups. Both MNE and CNE are linearly related to feed efficiency, thus representing promising parameters that need further attention in future studies.



Chapter 1

Protein metabolism in the rumen

Dietary protein digestion is characterised by complex degradation and neo-synthesis processes operated by the ruminal microbial population.

Protein is categorised in rumen-degradable protein (RDP) and rumen-undegradable protein (RUP). The RUP possesses the ability to pass through the rumen without undergo degradation. The RDP consists of true protein N and non-protein N (NPN):

the latter includes nucleic acids, amines, amides, ammonia, urea, nitrates, ammonium, amino acids (AA) and small peptides (Bach et al., 2005).

1.1 – Protein degradation

Dietary protein first undergoes a degradation process operated by extracellular proteases, through which peptides and AA are obtained: they are then carried inside bacterial cells, where peptides are further degraded to AA by intracellular peptidases. These AA can either be utilised to synthesise microbial protein, or undergo deamination, through which volatile fatty acids (VFA), CO2 and ammonia are obtained.

Several factors affect protein degradation: type of protein and solubility, ruminal passage rate, ruminal pH, substrate fermented (in particular, type and degradability of dietary carbohydrates), predominant bacterial species (Bach et al., 2005).

Simple proteins are categorised as follows according to their solubility level (Fantini, 1983):

1. Protamines 2. Histones 3. Albumins 4. Globulins 5. Glutelins 6. Prolamins 7. Scleroproteins 8. Phosphoproteins.



Albumins, globulins, glutelins and prolamins are the classes of major interest in ruminant nutrition. Albumins and globulins represent the main protein fraction of ingredients like alfalfa, as reported by Hood and Brunner (1976), and oat, linseed and sunflower (Wohlt et al., 1973). Whereas, glutelins and prolamins constitute the main protein fraction of corn, wheat and soybean meal, according to Wohlt et al. (1973). In the cited paper protein solubility was measured in common feedstuff:

solubility ranged from 3 to 93% and, in particular, they reported that feeds which major protein fractions were composed primarily of albumins and globulins were more soluble (32%-42%) than those composed primarily of prolamins and glutelins (25%-18%). According to Romagnolo et al. (1994), prolamins are more resistant to degradation than albumins, globulins and glutelins because of their AA composition:

prolamins are richer in neutral and non-polar AA, while albumins and globulins contain more basic and acidic AA.

Additionally, Yang and Russell (1992) reported that specific peptide bonds are more resistant to degradation than others: as they conducted an in vitro study to evaluate degradation of proline (Pro)-containing peptides, they observed that lysine (Lys)-Pro and methionine (Met)-Pro dipeptides were degraded much more slowly than Lys-alanine (Ala) and Met-Ala dipeptides.

Structural features of a protein condition its degradability: for example, proteins are degraded more slowly if disulphide bonds are present, or if they possess tertiary or quaternary structure, because of the presence of bonds within and between protein chains(Bach et al., 2005). Moreover, protein degradation is inversely related to ruminal passage rate, which can be controlled to reduce degradation by regulating the factors that influence retention time, such as feed amount and quality, particle size and forage to concentrate (F:C) ratio (Kamalak et al., 2005). Ruminal pH level affects protein degradation, which decreases as pH drops; however, it has been observed that protein degradation is influenced not only by pH level but also by substrate and microbial population, since diet composition can affect microbial population and its degradation abilities (Bach et al., 2005).

1.2 – Microbial protein synthesis

Bacteria synthesise microbial protein by using the AA obtained from peptide degradation operated by extracellular proteases and by intracellular peptidases, in



addition to ammonia N that generates from AA deamination; this ex novo-built protein is subsequently absorbed in the small intestine along with RUP.

In order to synthesise microbial protein, beside N sources, carbohydrates are necessary as energy source for the development of peptide bonds; in fact, one of the most important features to consider in order to promote microbial protein synthesis is the proportion between carbohydrates and N and their respective degradation rate, which is synchronisation. When dietary protein is digested more rapidly than carbohydrates are fermented, the energy supply is insufficient to synthesise microbial protein, leading to a great amount of N being lost as ammonia;

similarly, if protein degradation doesn’t occur rapidly enough in respect to carbohydrates fermentation, microbial protein synthesis decreases despite energy availability (Bach et al., 2005).

Beside synchronisation of ruminal fermentations, microbial protein synthesis is influenced by several factors: DMI, dietary N, F:C, and ruminal pH.

If DMI increases, microbial protein synthesis increases too; Pathak (2008) reported that the addition of straw and starch supplementation to barley-based diets, operated in order to increase DMI, significantly improved microbial protein synthesis. Concerning starch, according to Bach et al. (2005), readily fermentable carbohydrates such as starch and sugar promote bacterial growth more than cellulose does: they reported that infusions of increasing amounts of readily fermentable carbohydrates decreased ammonia N concentrations because of raised N uptake by ruminal microbes.

Dietary N amount must satisfy microbial requirements, but it also has to be balanced in terms of protein degradability: according to Pathak (2008), it seems that proteins characterised by lower rates of ruminal degradation tend to improve the efficiency of microbial protein synthesis, probably because of the better capture of released N by ruminal bacteria. Considering forage-based diets, the same author reported that 2 g of available N per 100 g of digestible organic matter is required for optimal microbial growth for animal fed forages.

The F:C influences both ruminal pH and microbial protein synthesis; the latter is reported to be low in cows fed high-concentrate diets but also in low-quality forage- based diets because of slow carbohydrate degradation (Pathak, 2008). As the proportion of forage increases, there is greater saliva flow, which determines a higher ruminal pH. Diets containing a mixture of forages and concentrates increase



the efficiency of microbial protein synthesis because of an improved rumen environment for the growth of different bacteria species (Pathak, 2008).



Chapter 2

Rumen fermentation synchronisation

The synchronisation concept indicates the supply of protein and energy, the latter as readily fermentable carbohydrates, in a way that allows ruminal microbes to utilise these elements simultaneously (Seo et al., 2010), thus optimising microbial protein synthesis and minimising N waste. Synchronisation is traditionally measured through the synchrony index (SI; Sinclair et al., 1993): a SI of 1 represents perfect synchrony between N and energy supply, while values between 0 and 1 indicate the degree of asynchrony (Piao et al., 2012).

2.1 – Protein and energy synchronisation

Seo et al. (2010) conducted a study on Holstein heifers with the aim to evaluate the effects of three diets characterised by an increasing SI (0.77, 0.81 e 0.83) on ruminal fermentations, N efficiency and microbial protein synthesis.

The animals fed the diet with the lowest SI showed higher ammonia N levels in the rumen and thus higher N excretion in urine, while the diet with the highest SI led to the highest level of microbial protein synthesis; the final results of the study showed that feeding a highly synchronised diet improved microbial protein synthesis and VFA production, while decreased total N excretion.

However, as it was outlined by the researchers themselves, there are several examples in literature of conflicting results according to the effects of synchronisation; anyway, they specified that they adopted diet formulation criteria aimed to minimise differences between the diets in terms of chemical composition and RDP concentration, regulating non-structural carbohydrates content in order to obtain the desired SI variations, while other researchers chose to work on other parameters, such as, for example, protein degradability.

An additional study was conducted by Seo et al. (2013) with the aim to evaluate the effects of the synchronisation of carbohydrates and protein supply on ruminal fermentations and microbial protein synthesis. They tested six feeds with different SI that contained non-treated corn or enzyme-treated corn as carbohydrate source and non-treated, enzyme-treated or formaldehyde-treated soybean meal as protein



source; feeds containing enzyme-treated carbohydrate or protein sources were characterised by the highest SI values (>0.8), while diets containing formaldehyde-treated protein presented the lowest values (0.71-0.72). The feeds had similar chemical composition, and this was not altered by the treatments applied.

For high SI feeds containing non-treated corn the highest DM digestibility value was recorded, while for feeds containing enzyme-treated corn a lower ammonia concentration was recorded (indicating a more efficient utilisation for microbial protein synthesis) and higher VFA values were obtained both in terms of overall production and individual VFA concentration, but it happened regardless of SI. The best values were obtained for enzyme-treated feeds, although they were characterised by an intermediate SI; this led the researchers to suggest that energy and protein supply availability could be the most limiting factor for ruminal fermentation and microbial protein synthesis, rather than the synchrony between the two nutrients.

Factors like nutrients availability, feed characteristics, type of N source are pointed out by Seo et al. (2010) as elements to work on in order to reach a practical application of synchronisation as a technique to improve animals’ productive efficiency.

2.2 – The Cornell Net Carbohydrate and Protein System

The Cornell Net Carbohydrate and Protein System (CNCPS) is a mathematical model that estimates cattle requirements and nutrient supply based on animal information, environment, and feed composition in different production systems. In order to predict animal requirements, different factors are considered: physiological state (lactation, pregnancy, and growth), body reserves and environmental effects.

The CNCPS uses feed carbohydrate, protein degradation and protein passage rates to predict extent of ruminal fermentation, microbial protein production, post-ruminal absorption, and total supply of metabolizable energy and protein to the animal.

This model has been successfully used to evaluate and formulate balanced rations for beef and dairy cattle (Fox et al., 2004); by inserting the inputs, it allows to estimate the accomplishment of animal requirements, while suggesting modifications in ration’s composition to increase productivity.



The model estimates protein and carbohydrate degradation from the degradation and flow rates of the various fractions and sub-fractions of dietary components.

Microbial protein yield and fermentation products are estimated by comparing carbohydrate degradation rate with protein degradation rate (schematised in Table 2).

Table 1: Crude protein and carbohydrate degradation classes (Fox et al., 2004; Tylutki et al., 2008).

CNCPS1 classes

Crude Protein Degradation Carbohydrates Degradation

A1 NPN2 Immediate Sugars Rapid

A2 True soluble protein (AA, small peptides, globulins, albumins)

Very rapid Starch Intermediate

B1 moderately degradable true proteins (prolamins, cell wall proteins, denatured proteins)

Intermediate Available cell wall


B2 Slowly degradable proteins bound to NDF

Slow - -

C Unavailable protein or protein bound to the lignin and tannins, Maillard protein

None Lignin None

1: Cornell Net Carbohydrate and Protein System; 2: Non-protein nitrogen.

The rates of passage of microbial products and undigested feed out of the rumen are also estimated with this system: the model assumes feed intake as constant, rate of passage and extent of digestion as a function of intake, and that all DMI is either digested or passed through the intestinal tract.

A CNCPS sub-model has been developed in order to predict dietary AA requirements: it estimates absorbed AA supply from predicted rumen microbial growth and composition as well as the amount and composition of RUP. The AA requirement estimation is based on milk yield and on the efficiency of conversion of AA digested in the duodenum to milk protein (Dinn et al., 1998).

Dinn et al. (1998) carried out an experiment aimed to test the possibility of reducing dietary CP to promote a more efficient N utilisation and its lower excretion in urine, but in a way that would not compromise total milk yield or true protein yield. To pursue their objectives, they used CNCPS to formulate balanced rations in terms of energy and protein supply and supplemented the diets with rumen-protected Met (RPM) and rumen-protected Lys (RPL).



They examined three diets characterised by decreasing CP percentage (18.3, 16.7 and 15.3 %), of which the last two were supplemented with RPL and RPM. As CP decreased, N efficiency increased while urine and blood urea N levels decreased;

milk yield decreased (from 34.2 to 32.8 Kg/d), but true protein yield was nearly unchanged. According to the researchers, the use of CNCPS for diet formulation and the supplementation of the latter with RPL and RPM were decisive for these results, since there are different variables to consider while balancing rations in order to meet the cow’s AA requirements.

Overall, the researchers outlined the advantages of formulating diets with the aim of improving N efficiency: reducing N excretion in urine and faeces results in smaller environmental impact, therefore permitting more sustainable farm systems.



Chapter 3

Nitrogen efficiency and protein nutrition in dairy cows

Dietary protein in a complete diet is expressed as CP, which is defined as the feed’s N content multiplied for a factor of 6.25. Nowadays, CP proportion in the diet of high- producing dairy cows raised in the Parmigiano-Reggiano cheese production area ranges approximately between 15% and 18% (Righi et al., 2007; Simoni et al., 2020).

For many years, CP percentage represented the main feature to work on while formulating diets, but during the last few decades a far more complex and articulate scheme took shape concerning protein nutrition. Thanks to the results of several studies, highest importance was assigned to the qualitative features of dietary protein and the main goal set was to meet the cow’s AA requirements (Schwab and Broderick, 2017). Along with AA profile balance, research stressed the importance of a balanced amount of protein and carbohydrates to synchronise ruminal fermentations, thus increasing microbial protein synthesis (Seo et al., 2013).

Regulating protein nutrition is the key to increase N use efficiency: it means optimising milk production and composition while reducing N waste in terms of urea, which is secreted in milk, faeces, and mainly in urine, thus significantly contributing to environmental pollution (Dijkstra et al., 2011; Foskolos and Moorby, 2018).

Table 1 shows a few examples of protein composition in diets formulated for high-producing dairy cows.

Table 2: Ranges of crude protein (CP), soluble protein, rumen degradable protein (RDP) and rumen undegradable protein (RUP) in diets of high-producing dairy cows.

Reynal and Broderick, 2005

Ipharraguerre and Clark, 2005

Bahrami-yekdangi et al., 2016

CP %DM1 17.2-18.8 14.8-19.0 14.8-16.4

Soluble CP (%CP) - 34.9-36.4 -

RDP (%DM) 7.7-12.5 9.5-11.9 9.3-10.9

RUP (%DM) 6.3-9.5 5.0-7.7 5.5

1: Dry matter.



3.1 – Rumen degradable and undegradable protein

The presence of RUP is necessary in a high producing dairy cow’s diet. However, some elements need to be considered in order to promote an efficient N utilisation:

this can be reached by regulating dietary RUP quality and its amount related to RDP, beside CP percentage and AA profile.

Bahrami-yekdangi et al. (2016) conducted an experiment on three groups of cows in early lactation fed three different diets characterised by the same RUP percentage (5.5% DM) but decreasing percentage of RDP (and so total CP too), supplemented with RPM; the aim of that study was to evaluate the effects of the CP and RDP reduction on milk production parameters and on N utilisation, given a constant RUP amount and a balanced Lys/Met ratio.

Milk yield and quality were not affected by RDP and CP amount; nevertheless, the decrease of RDP and total CP was accompanied by a significant improvement in the N efficiency with a reduction of N excreted in urine. The diet with the lower protein content (14.8 % CP and 9.3% RDP) resulted able to properly satisfy microbial RDP requests, keeping milk yield stable and permitting a more efficient N utilisation; beside these results, the RPM supplement has to be considered for its role in maintaining a balanced Lys/Met ratio by supplying limiting AA for milk production.

The balance between RDP and RUP amount is highlighted as main concept. An appropriate amount of RDP is in fact necessary to fully satisfy microbial requirements; a RDP deficit leads to insufficient ammonia production, and thus to a decrease in microbial protein synthesis, while a RDP excess leads in turn to an excessive ammonia production, which is absorbed and excreted as urea. Moreover, the RUP administered has to be easily digestible in the small intestine and present a balanced AA profile in order to promote milk production.

Similar concepts were analysed also by Sun et al. (2009) during a study were they compared four diets that differed for CP percentage (13.1% and 14%), RUP percentage (5.2% and 6%), RUP digestibility (lower and higher), and presence or absence of Met supplementation. Their aim was to verify if it was possible to maintain milk yield and composition while improving N efficiency by feeding less digestible RUP supplemented with Met.



The most satisfying results were obtained among the group fed 13.1% CP and 5.2% highly digestible RUP with Met supplementation: in this group the highest values in terms of milk fat and true protein percentage were obtained, together with the lowest milk urea N level and the lowest N excretion in urine. Moreover, milk yield was similar to the one obtained for the groups fed higher RUP percentage without Met supplementation. These results led the authors to conclude that RUP digestibility and AA profile balance may be more important than just the amount of RUP.

A study conducted by Wright et al. (1998) could be considered in support of the same conclusion: it involved three diets that differed for RUP percentage (4.5, 14.9, and 29.1% DMI), and which composition was studied to obtain an AA profile similar to bovine caseins for Met, Lys, histidine (His), phenylalanine (Phe) and threonine (Thr). It was observed that milk protein yield increased linearly as RUP percentage increased, but N efficiency was highest for the diet with the lowest RUP amount: in fact, a linear increase of N excretion in urine was recorded as RUP amount increased. Moreover, during the same study, restrictions were applied to feed intake: part of the basal component of the diet (RUP excluded) was reduced by 10 and 20%, thus limiting energy supply. As a consequence, this led to a higher N excretion in the urine of the cows on which the 20% feed intake restriction was applied: to explain this event the researchers considered a likely disequilibrium between protein and energy supply, the latter resulting insufficient for an efficient N utilization by ruminal microbes.

3.2 – Limiting amino acids

By the end of the 60’s some researchers started to study the effects of supplying certain AA on milk production; during the first experiments, which involved casein infusions in the abomasum, conducted on cows which diet was considered fulfilling in terms of protein supply, an increase of 12-13% in milk yield and its protein fraction was observed. Thus, since early 70’s different studies have been carried out with the aim to individuate the most limiting AA for milk protein synthesis (Schwab and Broderick, 2017).

The studies’ results show that Met and Lys are the most limiting AA in most nutritional situations: this is due to the fact that most feed proteins contain lower



amounts of Lys and Met, particularly of Lys, than microbial protein (Schwab, 1995), as shown in Table 3. It could then be necessary to supplement rations with the goal to administer AA sources able to resist ruminal degradation but also easily absorbable in the small intestine (Schwab and Broderick, 2017).

Table 3: Lysine, Methionine and essential amino acids (EAA) composition of ruminal bacteria compared to the main dietary protein sources (Schwab, 1995).

Lysine (%EAA) Methionine (%EAA) EAA (%CP1)

Bacteria 17.3 4.9 40.0

Alfalfa forage 11.1 3.8 40.7

Soybean meal 13.7 3.1 47.6

1: Crude protein.

Samuelson et al. (2001) conducted two experiments to evaluate the effects of supplementing RPM on milk parameters of early and mid-lactating Holstein dairy cows fed alfalfa hay-based diets. Experiment 1 involved early lactating cows, and the control diet was formulated to supply 104% of the estimated Met requirements according to CNCPS, while the experimental diet was supplied with 30 g/d of RPM, thus increasing predicted metabolizable Met to 139% of Met requirements; in experiment 2, which involved mid-lactating cows, the control diet was formulated to supply 124% of the estimated Met requirements, while the experimental diet was supplied with 15 g/d of RPM plus 15 g/d of DL-Met, with predicted metabolizable Met thus raised to 155% of Met requirements.

The RPM supplementation in the early lactating cows had the effect to increase milk yield and fat percentage, but had no effect on protein percentage; conversely, in experiment 2 it had no effect on fat percentage in the mid-lactating cows, but it increased protein yield and percentage: the researchers concluded that RPM supplementation may be useful to improve milk composition, but the effects depend on dosage, stage of lactation and type of supplemental Met.

It is a fact that certain AA limit milk protein synthesis; however, a balanced and synchronised diet allows to obtain satisfying production parameters by improving N utilisation and thus increasing AA flow to the duodenum. The supplementation of limiting AA could represent a refined practice in protein nutrition to pursue better



milk parameters, but it is first necessary to expand common knowledge on diet regulation in order to obtain the finest results possible.



Chapter 4

Nitrogen efficiency and environmental pollution

During the last few decades public opinion’s awareness has increased towards the effects of human activities on the environment; this issue is still widely discussed and analysed worldwide, and despite remarkable results have been obtained thanks to the agreements reached between many nations through important steps such as the Kyoto Protocol in 1997 and the Paris Agreement in 2015, environmental protection and in particular the fight against climate change are still matters of major concern, on which nowadays the attention is maintained high particularly thanks to the “Fridays for Future” movement.

This spirit of awareness involves dairy industry too: since different European countries recognise dairy farms as the main N pollution source (Castillo et al., 2000), it becomes necessary to adopt sustainable management procedures.

4.1 – The nitrogen cascade

The continuous spill of anthropogenic N in ecosystems has led to the N cascade phenomenon (Foskolos and Moorby, 2018), that is the accumulation of reactive N and its atmospheric and hydrologic dispersion; this implicates relevant consequences for human, animal and environmental health.

Reactive N includes all N compounds in Earth’s atmosphere and biosphere having biological, photochemical and radioactive activity, such as ammonia, ammonium, nitric oxide, nitrates, urea. Before the development of human life, reactive N formation from molecular N (N2) occurred through natural processes, as microbial N fixation, and was appropriately balanced by denitrification processes, which convert reactive N back to N2; as time went by, this balance went missing, and nowadays reactive N accumulates in the environment at all levels. In particular, since 1960 reactive N accumulation has accelerated sharply: during the last decades reactive N human production has exceeded the one operated by all the natural systems. This accumulation is destined to increase, together with human population and per capita resource utilisation (Galloway et al., 2003).



The N cascade consequences fall on human health and on ecosystems: reactive N increase in the atmosphere implicates a higher production of toxic aerosols which induce severe respiratory diseases, cancer and cardiac diseases in human beings, and N together with sulphur is responsible for water acidification and the consequent loss of biodiversity in lakes and watercourses. Reactive N is the main cause of pollution for coastal ecosystems, where it causes eutrophication, hypoxia, loss of biodiversity and habitat degradation; but most of all, it contributes to climate change and ozone layer depletion (Galloway et al., 2003), consequently impacting on all life on Earth.

4.2 – Dairy industry and nitrogen pollution

Concerning dairy industry, different management and nutritional expedients can be carried out to reduce N pollution. As reported by Castillo et al. (2000), ammonia levels in the atmosphere have raised by 50% since 1950, and one of the main causes of this change has been identified in the intensification of animal production systems, which implicates a greater number of animals per farm. The main factor involved in ammonia volatilisation is represented by urea concentration in urine: the researchers agree that an appropriate nutrition management can considerably reduce ammonia emissions.

Castillo et al. (2000), by examining different publications, realised a quantitative analysis of dairy industry’s contribution to N pollution, considering the relationship between ingested N and N excreted in bovine urine and faeces. In particular, while comparing diets containing 200 g CP/DM with diets containing 150 g CP/DM, they noticed for the latter a remarkable reduction of ureic N in faeces, where the level dropped by 21%, and most of all in urine, with 66% less of excreted urea. The N amount that corresponds to 150 g CP/DM, that is 400 g/d, seems to be the critical point above which N excretion in urine increases exponentially: thus, according to the researchers, keeping dietary CP amount below 150 g/DM effectively reduces ammonia production.

Overall, the researchers agree on nutritional strategies to monitor N emissions: first, reducing N intake, which has to be calibrated on the animal’s requirements; beside it, reducing instable N amount in the rumen by regulating protein degradation; then, improving the synchrony between energy and protein supply.



Chapter 5

Casein nitrogen as an efficiency parameter

Milk protein is divided in two main categories: caseins, which account for 80%, and soluble protein, consisting of the remaining 20%. Casein amount and composition is the most important characteristic for cheese making, as it defines milk’s possibility to be transformed into cheese (Stefanon et al., 2002). Considering that 43% of Italian milk is destined to PDO cheese production, and 16% is used to produce Parmigiano-Reggiano cheese (CLAL, 2018), maximising milk casein yield and quality assumes major importance for many dairy farmers.

Although milk protein composition is mainly controlled by genetics, diet composition can affect milk protein yield and casein fraction.

Milk protein is positively influenced by energy concentration and starch amount; it has been observed that supplying low rumen-degradable starch such as corn increases milk production and protein content, while dietary starch amount seems able to modify casein fractions by increasing αs1-casein and αs2-casein percentage (Stefanon et al., 2002). Auldist et al. (2000) also reported that grazing cows fed ad libitum forage produced significantly more α-casein, β-casein e κ-casein: they supposed that the higher energy intake could presumably have spared AA from gluconeogenesis, thus increasing the AA supply available for milk protein synthesis.

Quite a few studies have been carried out in order to compare different protein sources and their effect on milk protein production: for example, Broderick et al.

(2015) carried out a study to evaluate the effects of replacing soybean meal with canola meal. As they reported, evidence exists that CP from canola meal can increase milk protein yield more than soybean meal does. The researchers reported that beside improving feed intake and milk yield, the use of canola meal improved milk protein yield and N utilisation. Křížová et al. (2017) specifically reported effects on casein yield. They compared the effects of feeding Holstein dairy cows a diet containing extruded rapeseed cake with those of a diet where part of the extruded rapeseed cake was substituted with extruded full-fat soybean. The cows fed the diet containing soybean presented higher DMI and milk yield, but showed significantly lower values in terms of milk protein and casein content; in particular, casein content



amounted to 27.4 g/Kg of milk in the soybean + rapeseed diet, against the 29.0 g/Kg in the rapeseed-only diet.

Another study performed by Mordenti et al. (2007) focused on examining ingredients that could substitute soybean meal in diet formulation for Holstein dairy cows raised in the Parmigiano-Reggiano cheese production area. The researchers compared a diet containing 10% soybean meal with a diet where it was substituted with 10.1%

peas and 10.1% fava beans. As they compared the effects of both, they found that the cows fed the diet containing peas and fava beans had a depression of DMI (3.5% less) and milk yield (3.9% less), but significantly increased casein concentrations (2.52 ± 0.07%), compared to the soybean meal diet (2.49 ± 0.05%).

Further research on regulating milk casein through diet formulation and on different feed sources’ properties could be of valid application in the field of protein nutrition of high-yielding dairy cows which milk is widely used for cheese production, as it occurs in Northern Italy.



Chapter 6

Experimental contribution


Environmental pollution has nowadays become matter of greater public attention; it is suitable to remind that the European Commission founds projects and other initiatives in order to implement the application of the European Union policies concerning this issue (European Commission website, 2020). Specifically concerning N pollution, the accumulation of reactive N in the environment has been occurring considerably faster since a few decades ago (Galloway et al., 2003), impacting on human beings, animals and ecosystems and constituting a serious risk for their health (Foskolos and Moorby, 2018).

Many European countries recognise dairy farms as one of the main N pollution sources (Castillo et al., 2000); it is estimated that livestock ammonia emissions account for 39% of ammonia emissions worldwide (Zou et al., 2020). Veterinary science can give its contribution to the fight against environmental pollution by studying manners to improve dairy cows’ N efficiency, with the aim of maximising the amount of dietary N ingested by the cow that is transformed into milk protein N (Nadeau et al., 2007) while reducing N losses in the form of urea, which concentration in urine represents the main factor involved in ammonia volatilisation (Castillo et al., 2000). For a long time, while formulating diets, farmers tended to feed excess dietary N to the cows in the urge to avoid N shortage and assure the maximum milk production level, thus causing the overfeeding phenomenon (Chen et al., 2011). Nowadays several dynamic models are available (the CNCPS, for instance) which allow to evaluate ruminal activity in a very accurate way (Fox et al., 2004) and show that high milk production can be assured with less CP as long as the cow’s protein requirements are satisfied (Chen et al., 2011).

Urea excretion can be reduced through an accurate nutrition management, which involves reducing the amount of dietary CP, focusing on its qualitative features in order to meet the animals’ requirements in terms of AA and degradation rates, and calibrating protein and energy supplies to synchronise fermentations, thus promoting a more efficient N utilisation (Castillo et al., 2000).



The N efficiency is traditionally measured through the milk N:N intake ratio (Neal et al., 2014; Christensen et al., 2015); since cheesemaking represents an important branch in the market of dairy products in Northern Italy, considering as an example Parmigiano-Reggiano cheese production, it could be a matter of interest to evaluate N efficiency through milk casein yield. Thus, a new index to evaluate N use efficiency is proposed: CNE, defined as the amount of dietary N that is converted into caseins, and there is chance it could represent not only a way to evaluate N use efficiency but also a parameter to esteem cheese yield.

Therefore, the aim of this work is to evaluate CNE of dairy cows raised in the Parmigiano-Reggiano Consortium area and fed a hay-based TMR.



Materials and methods

This study was carried out in accordance with the Italian Legislation on animal care (DL 26 04/03/2014). Nineteen Italian Holstein dairy farms were randomly chosen within the Parmigiano-Reggiano cheese production area according to two criteria:

the exclusive breeding of Holstein cows - to reduce the variability related to production parameters - and the use of TMR, to allow an easier DMI quantification.

The farms selected ranged from 60 to 367 lactating dairy cows, with an average of 193.53 ± 93.75 cows and an average milk production level of about 33.49 Kg/cow/day.

For each farm (experimental unit), the sampling procedures were scheduled over 3 days. The sampling procedures took place in the early morning, collecting: on the first day, a sample of the TMR offered; on the second day, samples of the refused TMR first and of the offered one then; on the third day, a sample of the TMR refused, as shown in the following scheme:

Day 1 Day 2 Day 3

TMR refused TMR refused

TMR offered TMR offered

The TMR administered to the animals was sampled for two days and the total weight supplied was recorded; a day after all refusals were collected and weighed prior to wagon delivery. On every first day, the following information was recorded: number of cows, number of lactations, days in milk (DIM), average milk production and diet composition. The label of each feedstuff and the chemical composition of each forage included in the diet were collected.

Milk chemical sampling and analysis was performed weekly by the farms’ specific cheese factories during milk quality control inspections; for each farm the results of two analyses performed during two inspections were recorded, one of which took place before the sampling procedures, while another was performed after they were over.


22 Diet sampling and analyses

Diet samples were collected through a 1 Kg scoop. One Kg of TMR was collected from 10 points along the feedline and mixed in a handcart, then a subsample of 1 Kg was collected and stored in a sampling bag at – 20 °C. Then chemical analyses were carried out as described by Comino et al. (2015) to determine DM, humidity, NDF and starch content. Briefly, the feed samples were pre-dried at 55°C for 2 days and ground in a Cyclotec mill (Tecator Inc., Herndon, VA, USA) to pass 1 mm screen, then an aliquot of 5 g was dried at 103°C overnight for humidity determination, 0.5 g were used to perform the NDF determination and 2.5 g were used to analyse the starch content. The CP and N content was determined by Dumas method.

This method is used for N determination based on sample combustion at approximatively 900 °C in excess oxygen (Mihaljev et al., 2015). Solid samples are first weighed: a 200 mg sample is required for forages analysis, but the required weight ranges up to 500 mg depending on the ingredient examined. The samples are then wrapped in aluminium foil, compressed and placed inside the DUMATHERM ® (Gerhardt) through a transfer plate. They are then burnt, and the organic elements are oxidized. The combustion gases (O2, CO2, H2O, N2 and N oxides) are collected and passed through several traps; all gases are removed except N and N oxides. The bound N is transferred into molecular N and nitric oxides. The analysis gases are transferred with CO2 as a carrier gas through a catalytic post combustion zone onto a reduction zone. At this point, the conversion of the nitric oxides into N at hot tungsten takes place and the excess oxygen is bound. After a two-stage drying phase, the gas mixture flows to the thermo- conductivity detector via an electronic flow controller. A connected PC calculates the N concentration in the sample from the TCD signal of the N2 in the CO2 and from the sample weight. The content of CP is then calculated by multiplying the measured N amount by the appropriate factor of 6.25 and is expressed as a percentage (Mihaljev et al., 2015).

The average daily DMI was calculated subtracting the weight of the refusals to the weight of the TMR delivered, then dividing for the number of cows. The values of DMI obtained in the two consequent days were average to obtain the average DMI over two days.



Thus, knowing N and CP content, protein intake and N intake calculations based on DMI were performed.

Milk composition data

For each farm, the following data from two consecutive analysis were recorded:

fat %, total fat yield, total protein %, total protein yield, true protein %, true protein yield, casein %, casein yield, urea, somatic cells, lactose %, bacterial count.

Calculation of efficiency parameters

Energy corrected milk yield (ECM) was calculated according to the equation of Orth (1992):

𝑬𝑪𝑴 = [𝟎, 𝟑𝟐𝟕 × 𝒎𝒊𝒍𝒌 𝒚𝒊𝒆𝒍𝒅 (𝒌𝒈 𝒄𝒐𝒘/𝒅𝒂𝒚⁄ )]

+ [𝟏𝟐, 𝟗𝟓 × 𝒎𝒊𝒍𝒌 𝒕𝒐𝒕𝒂𝒍 𝒇𝒂𝒕 𝒚𝒊𝒆𝒍𝒅 (𝒌𝒈 𝑫𝑴⁄ )]

+ [𝟕, 𝟐 × 𝒎𝒊𝒍𝒌 𝒕𝒐𝒕𝒂𝒍 𝒑𝒓𝒐𝒕𝒆𝒊𝒏 𝒚𝒊𝒆𝒍𝒅 (𝒌𝒈 𝑫𝑴⁄ )]

The result was divided by DMI to calculate feed efficiency, and MNE was calculated using the following equation, as reported by Brito and Silva (2020):

𝑴𝑵𝑬 = 𝒎𝒊𝒍𝒌 𝑵 𝒚𝒊𝒆𝒍𝒅 𝑵 𝒊𝒏𝒕𝒂𝒌𝒆

Then, according to the same principle, CNE was calculated as follows:

𝑪𝑵𝑬 = 𝒎𝒊𝒍𝒌 𝒕𝒐𝒕𝒂𝒍 𝑵 𝒇𝒓𝒐𝒎 𝒄𝒂𝒔𝒆𝒊𝒏 𝒚𝒊𝒆𝒍𝒅 𝑵 𝒊𝒏𝒕𝒂𝒌𝒆

True protein N efficiency was calculated in a similar way, dividing milk total N from true protein by N intake.


24 Statistical analysis

A general statistical analysis was conducted to obtain an overall description of the farms involved in the study; the data were normally distributed. Then, the farms were grouped in two categories based on milk yield level (0 = ≤ 33 Kg/day; 1 = > 33 Kg/day). Group 0 included 9 farms while group 1 consisted of the other 10 farms.

The subsequent analyses were carried out comparing these two groups through the mixed model, with farm as random effect and productive level as fixed effect adjusted according to LSD test.

Statistical analyses were performed using the SPSS for Windows software package (SPSS Statistics for Windows, ver. 27. Chicago, IL, USA; SPSS, 2008).



Results and discussion

Population examined

The two groups considered were characterized by 213 and 176 average lactating dairy cows respectively for group 0 and 1. The average number of cows did not differ significantly between the two groups considered (P=0.588).

Diet composition

Diets were generally composed of alfalfa hay, mixed hay, fibrous and energetic feedstuff, corn meal, soybean meal, minerals and vitamins; further details on diets’

ingredients are shown in Table 4.

Table 5 shows the differences concerning the quantity of each ingredient administered to the two groups. Group 0 was averagely fed more forages and more fibrous feedstuff compared to group 1, who received way more feedstuff, especially of the energetic kind. Protein-based feedstuffs were more or less equally distributed, with group 0 receiving only a slightly higher amount. Sunflower was only used for group 0, while linseed and yeast were only included in group one’s diets. More mineral and vitamin supplement was used for group 1, which was also administered a greater amount of water.

Overall average F:C ratio (Table 4) was lower than the one requested by Parmigiano-Reggiano cheese production standards, which has to be no less than 1; however, when the two groups were confronted (Table 5), group 0’s average value was precisely 1, while group one’s ratio was lower: this could be due to the greater amount of feedstuff administered to group 1.



Table 4: Overall diet composition in terms of ingredients, forage and concentrate percentage and forage to concentrate (F:C) ratio.

Ingredients (Kg) Mean Minimum Maximum

Alfalfa hay 8.5 6.0 11.0

Mixed hay 5.3 1.9 15.5

Wheat 1.8 1.5 2.0

Straw 1.1 0.5 1.5

Feedstuff 5.8 0.5 13.5

Fibrous feedstuff 2.3 1.5 3.5

Beet pulp 1.5 1.5 1.5

Energetic feedstuff 6.2 2.0 11.5

Corn meal 5.7 3.0 10.5

Corn flakes 1.7 0.7 2.8

Sugar feed 1.1 0.2 2.5

Protein feedstuff 2.0 0.5 3.4

Soybean meal 1.3 0.4 2.5

Soybean flakes 1.7 0.4 4.1

Sunflower 1.1 1.1 1.1

Linseed 0.3 0.3 0.3

Yeast 0.1 0.0 0.2

Mineral-vitamin premix 0.5 0.0 3.0

Water 7.0 0.5 13.0

Forage % 46.9 33.1 57.8

Concentrate % 53.1 42.2 66.9

F:C 0.91 0.5 1.4



Table 5: Average amount of ingredients administered to each group, forage and concentrate percentage and forage to concentrate (FC) ratio. The minimum (Min.) and maximum (Max.) value for each item are also reported.

Group 0 Group 1

Ingredients (Kg) Mean Min. Max. Mean Min. Max.

Alfalfa hay 8.7 7.0 11.0 8.4 6.0 11.0

Mixed hay 5.4 1.9 15.5 5.2 2.3 8.8

Wheat 1.8 1.5 2.0 2.0 2.0 2.0

Straw 1.3 1.3 1.3 1.0 0.5 1.5

Feedstuff 3.5 0.5 6.0 8.2 4.3 13.5

Fibrous feedstuff 2.7 2.0 3.5 2.2 1.5 3.0

Beet pulp 1.5 1.5 1.5 0.0 0.0 0.0

Energetic feedstuff 2.6 2.0 3.2 7.7 3.6 11.5

Corn meal 5.9 4.1 10.5 5.3 3.0 6.5

Corn flakes 1.8 1.4 2.8 1.6 0.7 2.1

Sugar feed 1.1 0.2 2.5 1.0 0.6 1.5

Protein feedstuff 0.0 0.0 0.0 2.0 0.5 3.4

Soybean meal 1.5 0.4 2.5 1.0 0.5 1.4

Soybean flakes 1.7 0.4 4.1 0.0 0.0 0.0

Sunflower 1.1 1.1 1.1 0.0 0.0 0.0

Linseed 0.0 0.0 0.0 0.3 0.3 0.3

Yeast 0.0 0.0 0.1 0.2 0.2 0.2

Mineral-vitamin premix 0.4 0.0 1.5 0.5 0.0 3.0

Water 5.6 0.5 12.0 8.6 5.0 13.0

Forages % 48.7 33.1 57.8 45.3 39.7 51.6

Concentrate % 51.3 42.2 66.9 54.7 48.4 60.3

F:C 1.0 0.5 1.4 0.8 0.7 1.1



Results of general diet analyses are reported in Table 6: they are similar to those reported in other studies on Holstein dairy cows raised under the Parmigiano Reggiano Consortium (Comino et al., 2015; Righi et al., 2007; Simoni et al., 2020).

The DM content of the diets examined in this study was higher compared to the values obtained by Comino et al. (2015), while it was similar to those observed by Simoni et al., (2020). The average NDF content of our diets was higher than the values reported by the latter authors, while it was in accordance with the results obtained by Comino et al. (2015) and Righi et al. (2007). The average CP content observed in this study was in line with the values obtained by Comino et al. (2015), while Simoni et al. (2020) reported higher values. Our diets’ average starch content was similar to the values reported by Simoni et al. (2020), while Righi et al. (2020) reported lower values and Comino et al. (2015) observed higher values.

The diets were balanced to reach the minimum requirements for mid-lactating cows;

in fact, these diets were also similar in chemical composition to those reported by Ghelichkhan et al. (2018) and Christensen et al. (2015), who supplied diets that also included corn silage to Holstein mid-lactating cows.

Table 6: Overview of diet composition in the farms.

1: Dry matter; 2: Crude Protein; 3: Neutral detergent fibre.

Table 7 shows the mean values for the two groups. Diets administered to group 1 presented slightly higher values in terms of DM, CP and starch, but lower NDF, compared the ones fed to group 0. Despite group 0 received a higher amount of protein feedstuff, there was very little difference compared to group 1 (Table 5): diet analysis showed that the group 1 diet presented a higher CP percentage. The diet of group 1 had numerically higher starch content due to the higher amount of energetic feedstuff supplied to the animals; similarly, given that it was fed less

Mean Min. Max. Standard deviation

DM1 % 72.77 62.21 86.80 6.798

CP2 (% DM) 15.22 14.02 16.33 0.715

NDF3 (% DM) 38.29 32.73 42.98 3.329

Starch (% DM) 21.76 16.95 25.22 2.157



fibrous feedstuff than group 0, the latter’s diets’ NDF content appeared numerically higher.

No significant difference was found between the two groups, so it can be observed that overall diet composition appeared homogeneous.

Table 7: Average diet composition of each group.


Group 0 Group 1 SEM* P value**

DM1 % 72.32 72.53 1.582 0.946

CP2 (% DM) 15.15 15.28 0.164 0.686

NDF3 (% DM) 38.74 37.88 0.764 0.569

Starch (% DM) 21.50 21.98 0.495 0.626

1: Dry matter; 2: Crude Protein; 3: Neutral detergent fibre.

*SEM: standard error of the mean.

**P value: significance of the statistical test performed.

Intake levels

Overall intake values are shown in Table 8. The DMI values are similar to those observed in a study performed by Pacchioli et al. (2020) in a trial performed on Holstein dairy cows in the Po plain area in Northern Italy, even if the author reported a lower average milk production compared to the one reported in the present study.

The DMI was comparable to the values reported by Simoni et al. (2020) and Righi et al. (2007), which also reported similar farm production levels.

Comparable results in terms of DMI, CP intake and N intake were obtained feeding similar alfalfa hay-based diets in mid-lactating Holstein dairy cows (> 130 d) in other studies (Christensen et al. 2015; Ghelichkhan et al. 2018).



Table 8: Overall intake values.

Mean Min. Max. Standard deviation DM1 intake (Kg DM) 25.89 23.72 29.80 1.808

CP2 intake (Kg DM) 3.94 3.38 4.87 0.332

N3 intake (g DM) 630.15 541.42 778.50 53.123

1: Dry matter; 2: Crude Protein; 3: Nitrogen.

Average intake values were all found to be numerically higher in group 1 (Table 9), even if not significantly; this difference may be reflected in the higher average milk production that characterised the same group.

Table 9: Average intake values of each group.


Group 0 Group 1 SEM* P value**

DM1 intake (Kg DM) 25.31 26.31 0.415 0.404

CP2 intake (Kg DM) 3.85 4.02 0.076 0.173

N3 intake (g DM) 615.20 643.61 12.187 0.173

1: Dry matter; 2: Crude Protein; 3: Nitrogen.

*SEM: standard error of the mean.

**P value: significance of the statistical test performed.

Milk production and composition

Table 10 reports general milk yield and quality features. Average fat and protein percentages do not differ significantly from Italian Holstein standards; in particular, the average fat percentage observed is slightly lower than the one reported by ANAFI (National Italian Holstein and Jersey breeders association) in 2019 Holstein national statistics (3.67 against 3.81), while the average protein percentage is higher (3.38 against 3.36).

Fat, total protein, casein and urea values are comparable with those observed by Benedet et al. (2018) during a study where bulk milk composition was analysed in several Italian regions; although a certain variability was observed, given that herds



were from different regions, ad that herd composition was not completely homogenous because of the presence of minorities of other breeds (Brown Swiss and Simmental) in some herds, Holstein breed was still predominant among the herds, and the values obtained share the same range with this study.

Table 10: Overall milk composition.

Mean Min. Max. Standard deviation Average milk yield


33.49 28.29 42.20 3.331

Fat % 3.67 3.34 3.92 0.198

Total protein % 3.40 3.26 3.58 0.096

True protein % 3.38 3.24 3.56 0.097

Casein % 2.65 2.53 2.82 0.082

Urea mg/dl 22.98 20.00 29.10 2.447

Somatic cells x 1000/ml 202.48 92.00 356.07 82.568

Lactose % 4.89 4.77 5.01 0.059

Table 11 shows the results of the analyses performed on general milk data according to the two productivity levels. Obviously, the two groups differed significantly in terms of average milk yield, as it was the criterion adopted to create the two groups; however, milk composition did not differ significantly between them.

Concerning milk quality parameters, group 0 presented numerically higher fat, total protein, true protein and casein levels compared to group 1.

Group 0 presented numerically higher somatic cells count (SCC). It is known that SCC may affect the milk yield as reported by Ward and Schultz (1972). The latter reported a linear relationship between these two elements, and they used it in their study to determine milk loss due to mastitis in single quarters. Several studies has been performed on both pure and crossbreed Holstein cows using this relationship and observing an increase in milk losses in association with increased SCC (Boland et al. 2013; Hand et al. 2012). Moreover, group 1 was composed of averagely smaller farms: Sato et al. (2005) reported a trend for a lower SCC in small farms compared to bigger ones, thus it could be hypothesised that in smaller farms hygiene procedures were performed more easily and accurately because of the smaller number of animals.



Group 1 presented numerically higher urea secretion and slightly higher lactose levels compared to group 0. Group one’s higher urea secretion could be probably due to its diet’s numerically higher CP percentage and its numerically higher CP intake levels, two factors that might have determined a less efficient N utilisation with an increased urea formation and its subsequent secretion in milk.

Overall, no significant difference was found between the two groups’ values.

Table 11: Average milk composition of the two groups.


Group 0 Group 1 SEM* P value**

Average milk yield (kg/cow/day) 30.86 35.86 0.764 ≤ 0.001

Fat % 3.77 3.59 0.045 0.142

Total protein % 3.44 3.36 0.022 0.232

True protein % 3.41 3.34 0.022 0.231

Casein % 2.68 2.62 0.019 0.218

Urea mg/dl 22.80 23.14 0.561 0.815

Somatic cells x 1000/ml 233.23 174.80 18.94 0.215

Lactose % 4.89 4.90 0.013 0.765

*SEM: standard error of the mean.

**P value: significance of the statistical test performed.

Efficiency parameters

Results of efficiency parameters calculations are showed in Table 12. Feed efficiency values are in line with those obtained by Pacchioli et al. (2020) and by Ghelichkhan et al. (2018), who observed similar values in terms of MNE too.

Christensen et al. (2015) reported similar values in terms of ECM and MNE, while the feed efficiency values they obtained were higher than the ones observed in this study.



Table 12: Overall efficiency parameters.

Mean Min. Max. Standard deviation

ECM1 (Orth, 1992) 35.02 30.76 42.50 2.946

Feed efficiency 1.36 1.19 1.75 0.130

MNE2 0.28 0.24 0.37 0.032

True protein N efficiency 0.28 0.23 0.37 0.032

CNE3 0.22 0.19 0.29 0.024

1: Energy-corrected milk. 2: Milk nitrogen efficiency. 3: Casein nitrogen efficiency.

Efficiency parameters of each group are reported in Table 13. All the values were higher in group 1 compared to group 0; MNE, true protein N efficiency and CNE values did not differ significantly between the two groups, while significant difference was found on ECM and feed efficiency values. Concerning ECM, the difference exists since ECM is calculated on milk yield, and the two groups were formed precisely on the basis of their different milk yield; similarly, feed efficiency is a result of ECM divided by DMI, so the difference in terms of milk production influenced both the parameters.

Concerning MNE and CNE, the lack of difference between the two groups may be due to the low number of farms involved in the study and to the similarity between the two groups, which differed in terms of average daily milk production only by 5 Kg/cow/day (group 0: 30.86 Kg/cow/day; group 1: 35.86 Kg/cow/day).

Table 13: Average efficiency parameters of each group.


Group 0 Group 1 SEM* P value**

ECM1 (Orth, 1992) 32.77 37.04 0.676 ≤ 0.001

Feed efficiency 1.29 1.41 0.030 0.035

MNE2 0.27 0.30 0.007 0.164

True protein N efficiency 0.27 0.29 0.007 0.165

CNE3 0.21 0.23 0.006 0.166

1: Energy-corrected milk. 2: Milk nitrogen efficiency. 3: Casein nitrogen efficiency.

*SEM: standard error of the mean.

**P value: significance of the statistical test performed.


34 Casein nitrogen efficiency regressions

As CNE values of both groups were regressed on diet CP content, CP intake and N intake, correlation was found between these elements. Figure 1 shows the relationship between CNE and N intake: as N intake increased, CNE decreased, this mainly affecting group 1; the same trend was observed for dietary CP and CP intake.

The decreasing of CNE in response to increased levels of dietary CP, CP intake and N intake could be explained with a N overfeeding, which led to more consistent N losses. This observation is widely supported by the literature (Castillo et al., 2000;

Hristov et al., 2004; Nadeau et al., 2007; Chen et al., 2011), as many studies demonstrated that excess dietary CP results in decreased efficiency of conversion of dietary N into milk protein - thus affecting casein yield - and less efficient use of ruminal ammonia N for milk protein synthesis (Hristov et al., 2004).

Group 1 was more affected than group 0: this is probably due to the numerically higher average values in terms of dietary CP, CP intake and N intake presented by group 1. Moreover, group 1 presented a lower average F:C ratio: the higher amount of concentrate fed to the cows could have affected their N utilisation efficiency by lowering ruminal pH.

Figure 1: Casein nitrogen efficiency (CNE) regressed on nitrogen (N) intake level. Group 0 is represented by the colour blue, while orange represents group 1.

y = -0,0003x + 0,3972 R² = 0,4648

y = -0,0004x + 0,4588 R² = 0,6829

0 0,05 0,1 0,15 0,2 0,25 0,3 0,35

500 550 600 650 700 750 800


N intake (g DM)



Concerning N losses, milk urea levels were related to CNE (Figure 2). Group 1 presented a noticeable CNE decrease as milk urea levels rised, while group 0 didn’t respond such remarkably: again, this is probably affected by the numerically higher dietary CP, CP intake and N intake levels that characterised group 1, to which likely corresponded a higher N waste through formation of urea that was secreted in urine, faeces and milk, since a positive linear relationship exists between these elements, as reported by Castillo et al. (2000).

Figure 2: Casein nitrogen efficiency (CNE) regressed on milk urea. Group 0 is represented by the colour blue, while orange represents group 1.

A strong positive linear relationship was observed between CNE and feed efficiency (Figure 3): as the latter increased, CNE increased linearly, as was observed for MNE too. The overall N efficiency improvement could be explained by the higher capacity of converting feed into milk components (measured through the feed efficiency parameter), which automatically led to a higher capacity to convert ingested N into milk protein, and so, casein.

y = -0,0012x + 0,2391 R² = 0,0321 y = -0,0052x + 0,3497

R² = 0,2248

0 0,05 0,1 0,15 0,2 0,25 0,3 0,35

15 17 19 21 23 25 27 29 31


Milk urea (mg/100ml)



Figure 3: Casein nitrogen efficiency (CNE) regressed on feed efficiency. Group 0 is represented by the colour blue, while orange represents group 1.

y = 0,1705x - 0,009 R² = 0,7016

y = 0,1743x - 0,0163 R² = 0,8288

0 0,05 0,1 0,15 0,2 0,25 0,3 0,35

1 1,1 1,2 1,3 1,4 1,5 1,6 1,7 1,8


Feed efficiency




Gerelateerde onderwerpen :